new·Earn with mozg — 20% of every monthSend somebody here and take a fifth of every plan payment they make, for as long as they keep paying — not a bounty on the first invoice. Your handle is the link, the window is thirty days, and the commission lands on your balance the second they pay. Free to join: if you have signed in, you already have the link. mozg.sh/earnall news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

provider capabilities

6 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

DeepSeek model capabilities

DeepSeek model capabilities: `deepseek-v4-flash` supports text generation, object generation, tool usage, and tool streaming. `deepseek-v4-pro` supports text generation, object generation, tool usage, and tool streaming. `deepseek-v4-flash-vision-exp` supports text generation, object generation, image input, tool usage, and tool streaming. None of the standard models support image input except the vision variant.

DeepSeek reasoning streaming

DeepSeek V4 models support reasoning exposed through streaming. The stream yields parts with `type === 'reasoning'` for reasoning text and `type === 'text'` for the final answer.

ByteDance Seedream image models factory method

Create ByteDance image models using the .image() factory method on the byteDance provider instance. Models are referenced by their ModelArk model id (e.g. seedream-5-0-260128) or account-specific endpoint id (e.g. ep-...).

ByteDance Seedream image editing limitation

Seedream does not support mask-based inpainting. Edits must be described in the prompt, optionally with markers drawn on the input image for interactive editing, which is supported by dola-seedream-5-0-pro-260628.

ByteDance video models factory method

Create ByteDance video models using the .video() factory method on the byteDance provider instance.

ByteDance video supported aspect ratios

ByteDance video models support the following aspect ratios: 16:9, 4:3, 1:1, 3:4, 9:16, 21:9, and adaptive (image-to-video only). All models output MP4 video at 24 fps.

Give your agent this brain